VLE by Molecular Simulation

The metlrod requires suitable intennolecularpotentialenergy functionsZY(r) and solution of tire equations of statistical mechanics for the assemblies of molecules. As mentioned in Sec. 16.1, potential energy functions are as yet primarily empirical. Except for the simplest molecules, U r) caimot be predicted by ab initkr calculations, because of still-inadequate computer speed. Therefore, semi-empirical functions based on quantum-mechanical theory and experimental data are employed. 

Random displacement of molecules witluii each box. These are the usual moves of Monte Carlo simulation, insuring internal equilibrium and generating the ensemble upon wlucli die partition function is based, thus leading to a set of thermodynamic properties for the molecules of each box. 

Allen and D. J. Tildesley, ComputerSimulationof Liquids, Clarendon Press, Oxford, 1989 D. Frenkel and B. Smit, UnderstandingMolecularSimulations From AlgorithmstoApplications, Academic Press, San Diego, 1996. A. Z. Panagiotopoulos, Molecular Simulation, vol. 9, pp. 1-23, 1992. 

Gubbins, Applications of Molecular Tlieory to Phase Equilibrium Predictions in Models for Thermodynamic and Phase Equilibrium. Calculations, S. /. Sandler, ed., pp. 507-600, Marcel Dekker, Inc., New York, 1994. Meaning from the beginning, i.e., from first principles. 

Random transfer of molecules between tire two boxes. These moves alter compositions and chemical potentials /ij of tire species in tire boxes, ultimately bringing about equality of tire chemical potentials for each species in tire two boxes. These moves also contribute to tire evolntion of tire tliennodynamic properties of tire molecules in tire boxes. 

---